The primary goal of this project is to identify and examine how adults in the U.S. engage in distinct patterns of health behaviors, such as cigarette smoking, alcohol use, physical activity, preventive services, and sleep duration. We suggest that health behaviors are not necessarily one-dimensional or concordant (i.e., they undertake all healthy behaviors or all unhealthy behaviors), but rather, health behaviors may form distinct discordant patterns (i.e., they undertake a mixture of healthy and unhealthy behaviors). Our intent is to understand: how race/ethnic and sex groups are distributed across concordant or discordant patterns of health behaviors, if education patterns mediate these relationships, and whether health behavior patterns are associated with specific mental and physical health outcomes including psychological distress, general health, body mass, and cardiovascular conditions. We use latent class analysis, 22 measures of health behaviors from the 2004-2009 National Health Interview Survey (N=159,934), and social and human capital theories of health lifestyles to explore predictors of health behavior patterns. We expand upon previous research by applying sophisticated techniques to a large nationally representative sample in an attempt to explain race/ethnic and sex disparities in health behaviors and outcomes, and provide for targeted multi-behavior intervention strategies within the United States.